Dataset DOI: 10.5061/dryad.h44j0zpxc

Description of the data and file structure

We exposed male guppies to an experimental heatwave (32°C) or a control temperature (26°C) for 5 days. To assess the effects of heatwaves on cognition, we tested, in one group of males, memory and learning ability during repeated trials in a mate-search maze over four days. In another group of males, we tested performance in a suite of single-trial cognitive tasks, encompassing mate choice, inhibitory control, anti-predator responses, and habituation, over two days.

Files and variables

File: DatasetsDetails.rtf

Description: This text document contains a description of all the variables in the datasets.

File: DetourTestData.csv

Description: The detour test was used to evaluate males’ inhibitory control, specifically their ability to navigate around (i.e., ‘detour’) a transparent barrier to reach a resource located behind it.

Variables

• LineID: ID of the row (for sorting the dataset)
• MaleID: identification number of the focal fish (always a male)
• Block: Experimental block (the group of males undergoing treatment and tests simultaneously)
• Treatment: Temperature treatment (heatwave or control)
• LatencyOut: time until the fish left the shelter, in seconds
• CorrectTrial: 1 when fish went straight around the barrier to the goal (without ever touching the barrier), 0 when the fish went first to the barrier
• SolvedTrial: 0 when fish never went around the barrier to reach the goal, 1 when fish did solve (go around the barrier)
• TimeBarrier: time spent in front of the barrier, in seconds
• TotalTime: total time passed before reaching the goal, in seconds (includes time not in front of the barrier)

File: HabituationTestData.csv

Description: A habituation test was conducted to assess the male’s response to a repeated, non-harmful stimulus (acoustic and vibrational). This test evaluated the fish’s simple learning ability, specifically their capacity to learn to ignore (i.e., habituate to) a persistent yet inconsequential stimulus. After 5-minutes acclimation period, there was a 5-minute period of free swimming, and then the fish was exposed for 10 minutes to a constant, standardized disturbance stimulus repeated every 7 seconds. The trial was divided into three different periods: 5 minutes before the stimulus exposure (‘before disturbance’), the first 5 minutes after the onset of the disturbance stimulus (‘short disturbance’), and the last 5 minutes during stimulus exposure (‘long disturbance’).

Variables

• LineID: ID of the row (for sorting the dataset)
• MaleID: identification number of the focal fish (always a male)
• Block: Experimental block (the group of males undergoing treatment and tests simultaneously)
• Treatment: Temperature treatment (heatwave or control)
• Stimulus: no_before - 5 minutes before stimulus exposure (before disturbance). yes_1_short - first 5 minutes during stimulus (short disturbance). yes_2_long - second 5 minutes during the stimulus (long disturbance).
• SwimSpeed: average swimming speed while moving, in mm per sec
• TimeFreeze: nr of seconds spent freezing (during each 5 minute stimulus period)
• TimeMove: nr of seconds spent moving (during each 5 minute stimulus period)
• ORLE: observation-level random effect (to correct a model over overdispersion)

File: MateChoiceData.csv

Description: The mate choice test evaluated male mate choice. Given that mate choice is a cognitively demanding process involving the assessment and comparison of potential mates, we hypothesise that heat stress could impair the cognitive functions underlying these behaviours, leading to less strong or more inconsistent mate preferences.

(NA row in the dataset when male movements could not be tracked in a video due to a technical failure)

Variables

• LineID: ID of the row (for sorting the dataset)
• MaleID: identification number of the focal fish (always a male)
• Block: Experimental block (the group of males undergoing treatment and tests simultaneously)
• Treatment: Temperature treatment (heatwave or control)
• ChoiceTime: Total time (in seconds) spent in the female choice zones (combining the time with each of the two females)
• NochoiceTime: Total time spent in the no-choice zone (in seconds).
• TimeWithPreferred: Total time (in seconds) spent with the preferred female (the female with whom the male spent most of his choice time)
• TimeWithNonPreferred: Total time (in seconds) spent with the not preferred female (the female with whom the male spent least of his choice time)
• NrSwitches: The number of switches between the two females while in the choice zone.
• SwitchRate: Per minute rate of switching ((NrSwitches/ChoiceTime) * 60)
• TimeWithBig: Total time (in seconds) spent with the bigger female
• TimeWithSmall: Total time (in seconds) spent with the smaller female
• BigFemSize: size (in mm) of the bigger female (NA: measurement error)
• SmallFemSize: size (in mm) of the smaller female (NA: measurement error)
• FemDiff: size difference (in mm) between the two females
• BigFemPos: position of the bigger female’s compartment in the setup (randomised)
• OLRE: observation-level random effect (to correct a model over overdispersion)

File: MazeData.csv

Description: To assess whether exposure to a heatwave affects spatial cognition and learning in male guppies, we tested their performance in a maze using an associative learning protocol. In this test, the reward for successfully completing the maze was access to an unfamiliar virgin female. Each male was tested during 9 trials.

Variables

• LineID: ID of the row (for sorting the dataset)
• MaleID: identification number of the focal fish (always a male)
• Block: Experimental block (the group of males undergoing treatment and tests simultaneously)
• Treatment: Temperature treatment (Heatwave or Control)
• TrialNumber: trial number 1 to 9 in the maze (for each male)
• TestDay: test day 1 to 3 on which trial was done (3 trials per test day)
• TestTime: time in the day that a trial was conducted
• SolvedTrial: whether the fish completed/solved the maze (i.e. autonomously reached the other end of the maze within 10 minutes); 1 - solved, 0 not solved
• SearchTime: time taken to solve the maze (in seconds), maximum time of 10 mins (600 second) is assigned to uncompleted maze trials
• DeadEnd: number of times the fish entered a maze corridor with a dead end (NA - could not be estimated due to low movement)
• SwimBack: number of times the fish swam backwards into at previous corridor (in the direction of the starting point) (NA - could not be estimated due to low movement)
• NrErrors: total nr of errors: number of time swimming into a dead end plus times swimming backwards (NA - could not be estimated due to low movement)
• OLRE: observation-level random effect (to correct a model over overdispersion)

File: OpenFieldData.csv

Description: An open field test was performed to assess males’ activity and exploratory behaviour in a novel environment, and their response to a simulated predator risk. After a 5-minute acclimation period (not analysed) there was a 10-minute period of free swimming. After this, a conspecific alarm cue - simulating the presence of a potential predator - was added to the centre of tank, and the fish was observed for another 10 minutes.

(NA row in the dataset when male movements could not be tracked in a video due to a technical failure)

Variables

• LineID: ID of the row (for sorting the dataset)
• MaleID: identification number of the focal fish (always a male)
• Block: Experimental block (the group of males undergoing treatment and tests simultaneously)
• Treatment: Temperature treatment (heatwave or control)
• Stimulus: no_before = 10 minute period before alarm cue was added. yes_after = 10 minute period in presence of the alarm cue
• SwimSpeed: Average swimming speed (while moving) in mm per sec
• TimeFreeze: Time (in seconds) the fish spent freezing
• TimeMove: Time (in seconds) the fish was not freezing
• SpaceUsed: Nr of areas in the open field in which the fish was present at least once
• SpaceNotUsed: Nr of areas in the open field in which the fish never visited
• OLRE: observation-level random effect (to correct a model over overdispersion)

Code/software

The data can be viewed using the open sofware R. For our study, the data was used to investigate the effects of a heatwave on the cognitive performance of male guppies, using both linear and generalized linear mixed-effects models (LMMs and GLMMs) with the package lme4 in R version 1.4.0.

Access information

Other publicly accessible locations of the data:

• NA

Data was derived from the following sources:

• NA